Can I add a second Membrane to my Spectra Cape Horn Watermaker?

[Kenomac]

Just go on the Spectra website and download the pdf file.

[Frode]

Since I'm studying the potential effect of manipulating feed pressure, flow, and membrane surface myself, I was pondering a question similar to that of OP. Being an engineer, I tend to study the hell out of everything, then come up with a theory, and finally test the theory to see if it holds. I have an existing system with three 40" membranes in series, so my testing will go in the opposite direction, of reducing flow and/or pressure to observe the effect.

All things being equal, the higher the salinity, the lower the production rate. The higher the pressure, the higher the production rate. The higher the flow rate, the lower the salinity (farther down the membrane) - hence, higher production rate. The larger the membrane surface, the higher the production rate.

You can compensate for higher salinity with higher pressure, and get the same production rate as with lower salinity and lower pressure.

The OP was wondering if adding an additional membrane would increase the production rate. The answer is yes, it will, with a qualifier. If the flow rate is too low, causing the salinity in a significant portion of the second membrane to increase beyond the pressure/salinity threshold of the membrane, the production increase will be limited.

Chances are, operating the Cape Horn with a single feed pump would have that effect. However, also running the second feed pump, causing higher operating pressure on the membrane, as well as higher flow rate, should increase the production rate significantly (possibly around 50%). Since the Clark pump in this system is already dimensioned to handle two feed pumps, you will not place any additional stress on it by adding a second membrane.

Additional info for those interested in tech specs:
Documentation from Spectra/Katadyn indicates they have four versions of the Clark pump. The 10% version, used in models from the Ventura 200T through the Newport 400; the 7% version used in the Ventura 150; the 15% version used in the Newport 700; and the 20% version used in the Newport 1000.

These models vary primarily by the dimension of the piston rod, presumably to allow operation under higher pressure.

[Tellie]

I'm agin it.

[SV THIRD DAY]

I have to admit finding entertainment in these type of threads where experts in the field are ignored and dismissed for the fun of "what ifs". Of course if any of these what ifs made technical and economic sense they would have been done by the company in the first place. But in this Fake News World an internet chat room expert is given just as much credibility as the owners/designers/technical experts of the company. Kind of humorous when viewed in that light. But then again, I'm hauled out in the boat yard prepping for a mast pull and refurbishing before heading South to Mexico....so I'm trying to find humor where I can these days.

[Frode]

Not sure what you're trying to say there, Rich. I would have thought you, of all people, would have been highly supportive of not just trying out the configuration the OP was wondering about, but actually confirming the theory behind it.

Your own RO systems appear to be configured just the same way. In fact, my guesstimate of an additional 50% product increase was taken directly from your own writings on the subject.

I realize it's weird for a newbie like me to come in and question (neither ignoring nor dismissing) the existing forum experts' opinions. I ain't new to this subject matter, however, so don't let my low post count fool you.

Besides, and this is probably the most important fact (yup, not fake news) I will claim: If it weren't for those of us who constantly question "conventional" wisdom, and are more than ready to foot the bill of buying equipment, and setting up new configurations, just to experiment, no new ideas would ever come to fruition.

Spectra's in this business to make money. The only configurations they will support are the ones that make them more money. If you could enhance their $7,500 Cape Horn, by adding a $1,000 modification, and make as much water as their $13,500 Newport 400, they would certainly never admit to it.

[JstaRebel]

Originally Posted by Frode

Since I'm studying the potential effect of manipulating feed pressure, flow, and membrane surface myself, I was pondering a question similar to that of OP. Being an engineer, I tend to study the hell out of everything, then come up with a theory, and finally test the theory to see if it holds. I have an existing system with three 40" membranes in series, so my testing will go in the opposite direction, of reducing flow and/or pressure to observe the effect.

All things being equal, the higher the salinity, the lower the production rate. The higher the pressure, the higher the production rate. The higher the flow rate, the lower the salinity (farther down the membrane) - hence, higher production rate. The larger the membrane surface, the higher the production rate.

You can compensate for higher salinity with higher pressure, and get the same production rate as with lower salinity and lower pressure.

The OP was wondering if adding an additional membrane would increase the production rate. The answer is yes, it will, with a qualifier. If the flow rate is too low, causing the salinity in a significant portion of the second membrane to increase beyond the pressure/salinity threshold of the membrane, the production increase will be limited.

Chances are, operating the Cape Horn with a single feed pump would have that effect. However, also running the second feed pump, causing higher operating pressure on the membrane, as well as higher flow rate, should increase the production rate significantly (possibly around 50%). Since the Clark pump in this system is already dimensioned to handle two feed pumps, you will not place any additional stress on it by adding a second membrane.

Additional info for those interested in tech specs:
Documentation from Spectra/Katadyn indicates they have four versions of the Clark pump. The 10% version, used in models from the Ventura 200T through the Newport 400; the 7% version used in the Ventura 150; the 15% version used in the Newport 700; and the 20% version used in the Newport 1000.

These models vary primarily by the dimension of the piston rod, presumably to allow operation under higher pressure.

After reading your post and since you are doing studies using your SWRO, might I suggest that you study osmotic pressures and how to do proper calculations to figure them correctly. One other suggestion is to have a look at FilmTec website and the acceptable feed water requirements as well as acceptable recoveries with your specific membrane size. The issues that you were discussing in your post can be understood once you understand basic membrane dynamics. Hope your tests are successful.

[SV THIRD DAY]

First don't make the mistake in assuming that the chronology of a post reflect a direct relationship. I was writing and cracking myself up while typing, before your post hit the screen. So it wasn't targeted at you specifically. What I was commented on (and remember I've been here for years and have seen and commented on dozens of simular posts) was the lack of technical understanding if how a Spectra water maker vs a more conventional non-energy recovery type water maker that uses a simple back pressure regulating valve actually works. Trying to apply the concepts across different designs is a mistake so common and honestly so frequently made on these type of discussion boards, that a lot of times us experts in the field honestly get a little flippant in answering them over and over again. Perhaps we shouldn't and we each time the question is presented we should cut and paste the appropriate deep technical explanations in but we're only human and sometimes we get tired of answering the same questions over and over.

It's not that we get tired of educating people, I'm always happy to do that, and I help Cruisers build their own watermakers all the time and I'm happy to do so whether they buy anything from me or not. I'm always saving them from making the common mistakes of not matching the membrane surface area to their pump, the motor size to their pump, basically correcting all the errors and half-truths they read on the internet. The problem is when I'm being told I'm wrong by someone who hasn't spent the thousands or should I say tens of thousands of dollars in R&D to gather the scientific data, but is an expert on the internet none the less...ya...I can get a little flippant...hey everyone has faults.

Your attitude about Spectra only being in it for the money is part of the problem I see all the time. All companies are only in it for the money that goes without saying but what you're also leaving out is that for a company to be a successful business and make money they must provide a product that the consumer is willing to trade their money for. Spectra isn't in business to screw customers, as the current Bernie Sanders memes imply, Spectra is successful because they make a damn good water maker and customers like it and their service. So when the number one Spectra rep in the world is telling you something won't work, it's not because he wants to screw you and risk his reputation and keep you from learning about some magic shortcut. It's because it honestly isn't cost-effective. They've spent millions in research and development and if they could more cheaply bring a higher output unit to the market of course they would it would cripple the competition and I would retire. But with the laws of physics being what they are along with current membrane technology, I have plenty of good years left in me before I head to that Giant Taco Stand in the sky.

Oh and one other thing... if anyone's learned anything about me by now its that half of my post are intended to be humorous with a little good-natured smart ass thrown in just for fun.

Off to High School Graduation for my youngest son...then Mexico here we come again.

Cheers

[Larry Florida]

Originally Posted by Frode

Since I'm studying the potential effect of manipulating feed pressure, flow, and membrane surface myself, I was pondering a question similar to that of OP. Being an engineer, I tend to study the hell out of everything, then come up with a theory, and finally test the theory to see if it holds. I have an existing system with three 40" membranes in series, so my testing will go in the opposite direction, of reducing flow and/or pressure to observe the effect.

All things being equal, the higher the salinity, the lower the production rate. The higher the pressure, the higher the production rate. The higher the flow rate, the lower the salinity (farther down the membrane) - hence, higher production rate. The larger the membrane surface, the higher the production rate.

You can compensate for higher salinity with higher pressure, and get the same production rate as with lower salinity and lower pressure.

The OP was wondering if adding an additional membrane would increase the production rate. The answer is yes, it will, with a qualifier. If the flow rate is too low, causing the salinity in a significant portion of the second membrane to increase beyond the pressure/salinity threshold of the membrane, the production increase will be limited.

Chances are, operating the Cape Horn with a single feed pump would have that effect. However, also running the second feed pump, causing higher operating pressure on the membrane, as well as higher flow rate, should increase the production rate significantly (possibly around 50%). Since the Clark pump in this system is already dimensioned to handle two feed pumps, you will not place any additional stress on it by adding a second membrane.

Additional info for those interested in tech specs:
Documentation from Spectra/Katadyn indicates they have four versions of the Clark pump. The 10% version, used in models from the Ventura 200T through the Newport 400; the 7% version used in the Ventura 150; the 15% version used in the Newport 700; and the 20% version used in the Newport 1000.

These models vary primarily by the dimension of the piston rod, presumably to allow operation under higher pressure.

G'Day Frode.... it's a Norwegianname yes?.

I'm in the industrial membrane business 35 years. Have designed, installed and started up large industrial systems. I'm in agreement with your post.

In my experience membrane systems are deceptively simple.

[Frode]

Originally Posted by Larry Florida

G'Day Frode.... it's a Norwegianname yes?.

Aye, Larry, it is. My mother didn't predict that I would one day emigrate to the US, so didn't bestow upon me a more internationally recognizable name. But, I'm having fun with it. Tolkien's Frodo and I share the same Norse name root.

I certainly don't want to second-guess people like Tellie, with his huge amount of experience. I don't have the specs for the Cape Horn, so it's possible it can't create the necessary pressure and flow rate for the second membrane to be productive. Until the specs are proven not to support OP's desired configuration, I think it is important to keep an open mind about the possibility.

Spectra has done a great job, and their pricing model probably makes lots of sense. I am under absolutely no impression they're trying to screw anyone - they're simply charging what people are willing to pay. They're in the business to make money, and this is not a high volume business.

I love their core system, but am less taken by all the automatics. Also, having combined HW and SW skills, I can custom build all that control stuff for a tiny fraction of what they charge. But, I don't have to provide customer support for it, and I don't have to make a product that's idiot proof (nothing on my boat is), which they do. Again, fully supportive of their pricing model making sense. From what I have heard, all Spectra owners are very happy with the support they get, and that accounts for a lot. Not something I desire, but good for other people.

I have what used to be considered one of the better RO systems available at the time, an 800 gpd Filtration Concept system, with three 2540 membranes. All the components are serviceable, but I want the Clark pump as the primary high pressure device (for noise level more than power consumption). However, I don't want it enough to shell out the $16k for a similarly rated Spectra system, so I need to adopt, adapt, and improve. Hence, my deep interest in this subject, and loads of research to help me have an educated approach.

[Frode]

Oh, and maybe this isn't everyone's way of looking at it, but at $180 per membrane, I could change them out each year, and after ten years of cruising, be massively ahead with my less-than-ideal, faster-fouling, re-configured system than replacing it with another, higher capacity commercial solution.

[SV THIRD DAY]

It's not the membrane surface area limiting you....its the Hp sea water flow. Flow at pressure costs energy or mucho dierno from an energy recovery pump. As you starve the membrane for flow its just not a membrane life issue but your TDS also increases otherwise we would always want to add another membrane in series, as the cheap and easy way to increase production. As Scotty said, "you can't break the laws of physics".

[Pizzazz]

Let's break down the original question into it's components.

1) Can you increase watermaker output by adding more membranes (effectively increasing the product recovery rate)? Emphatic Yes, as long as you keep the pressure above 800 psi. The membranes reach their highest efficiency at 8% recovery but as Dow says in the spec sheet, recreational cruisers run them successfully at up to 20-30% recovery rate. The membranes will last less than three years vs. 10 years but at a cost of $189 per membrane it is a moot point. This has been discussed many times and you can assume it is a fact (or talk to Dow).

2) Will this work with the Clark pump? Most likely yes. We know that Spectra watermakers are very well tuned for high efficiency, achieving 1.1A per gallon (I believe the reverse osmosis energy limit is around 0.78A per gallon or lower, search the web). Most likely this high efficiency is achieved through tuning, the pressure, the flow rate and the pump. If the product recovery is around 8-10% as stated here, you can easily double this with just adding a second membrane. The efficiency of the system may go down (you may go up to 2.0A/gallon for example) which will defeat the purpose (better to run your current tuned watermaker longer) and the membrane life will shorten.

My guess is that the efficiency will go down only a little, so you will benefit from the second membrane.

[Pizzazz]

A few more points:

The problem with selling higher capacity watermakers is that the cost does not change significantly with the size of the watermaker. Motors cost $400 independent of hp, the high pressure pump is $900 independent of capacity and each additional membrane + housing is $600. If you add the fittings, hoses and gauges, you can build a 12 gph watermaker for $2,500 and a 24 gph version for $3,000. The rest is knowledge, time and profit. In case of Spectra we need to add a premium for the Clark intensifier and the relevant patents. So, a lot of this talk about running membranes at lower recovery rates comes from vendors trying to justify higher prices for high capacity watermakers.

Also, think about how you use your watermaker. If going on a long-term cruise far from civilization (may be less than 5% of the members here) and you depend on the watermaker for drinking quality water then you should spend $10,000 for the best system on the market. However, the majority of us could be served by a cheaper, less reliable watermaker. In my case, I bring bottled water for drinking and cooking (as I do at home) and make water for washing, bathing, cleaning, etc. Not a big deal if hoses are not food grade. Not a big deal if the system breaks down while on a three day trip to Catalina. We will survive (it has not broken yet

Someone asked about the energy requirements of the manual survivor. It is 50 watts or 12.5V x 4 amps. I have successfully used a windshield wiper motor for $69 to power one.

Someone mentioned 320 ppm output - you should consider changing the membrane, I aim to replace membranes when I go above 200 ppm. You either run oily or clorinated water to damage the membrane. When I experimented with the Survivor 35 watermaker, I was producing 150 ppm water with a membrane that spent 20 years in dry storage. These things are pretty much bullet proof unless you damage them deliberately. Good luck.

[SV THIRD DAY]

Originally Posted by Pizzazz

1) Can you increase watermaker output by adding more membranes (effectively increasing the product recovery rate)? Emphatic Yes, as long as you keep the pressure above 800 psi.

Simply Not a True Statement.
Sorry...but this is exactly how internet/chat room rumors get started, so lets quash this one quickly before someone blows their hard earned cruising cash on a bogus concept. It's this type of "Emphatic Yes" comments that I deal with all the time when folks call me up wanting to buy water maker part to either upgrade their current system or when building one from scratch. This isn't marketing hype or water maker chat room theory 101, but rather based on real life testing data.

If Pizzazz's statement was true, then I would be able to add a 3rd Membrane to our SM30 Water maker and increase the production, as long as I run the system at 800psi. So why don't I do that? The answer is because the 3rd membrane would not have enough flow through it and that doesn't just cause early membrane death (which as noted at $190ea that could be lived with) but it would also dramatically increase the product water TDS to the point that the 3 membrane system would not meet the 500ppm safe drinking water limits that is pretty standard in the water maker world.

Ok, so lets then add a second membrane to a one membrane system if you think my example of a 3 membrane system is on the extreme. Heck...that's the only difference between the Cruise RO SM20 and SM30 right...lets just add a second RO Membrane to any water maker and Sha-zam. Ah....no...sorry. The reason it works in our case (and others like Echotec this isn't rocket science) is because we are not just playing with the membrane surface area but also paying attention to the high pressure sea water flow rate. So to understand this better, what are the things that control fresh water production?
1. Membrane Surface Area
2. Sea Water Flow Rate
3. Membrane Operating Pressure
4. Sea Water Temperature
5. Sea Water TDS

Now without going to crazy extremes of heating up your raw sea water inlet stream, you have 1,2 and 3 that you can control and manipulate and then 4 and 5 are just what they are, BUT CAUTION you can't ignore Nos 4 and 5 or your water make can crap out when going from the normal 68-70-deg water maker rating temps up to temps that can be found in the Tropics. I've spent many summers at close to 90-deg water temps in the Sea of Cortez.

Data...lets looks at some data from a 40" Dow membrane, 68-degs F, in San Diego Bay, at 800psi. Membrane 2 (and 3) plumbed in series.

Hp Sea Water Inlet Flow/Membrane 1 Product /Membrane 2 Product
0.5GPM / 8GPH / NA
0.8GPM / 12GPH / NA
1.6GPM / 21GPH / 13GPH (our SM30 unit)
2.3GPM / 24GPH / 16GPH (our SM40 unit)
4.2GPM 25GPH 25GPH
4.2GPM / 25GPH / 25GPH Membrane No 3 10GPH (our SM60)

You can see that the product water production rate is not linear to sea water flow rate. The reason the 0.8GPM and 0.5GPM pump data set has an NA for membrane two was because the TDS was off the charts...IE...the membrane dropped below the minimum flow rate for not just long membrane life but for it to function. THIS is what people are not taking into account. It doesn't matter if you are willing to take a decrease in membrane life in exchange for extra production once you drop below the minimum flow rate. This isn't theoretical...it's real life and why simply telling folks they can always add a membrane and increase their product water production is BOGUS if their water maker was already nearing the sea water flow rate cut off point. Remember sea water flow rate at 800PSI costs you AMPS, so most water makers are already playing the game of minimizing flow rate as much as possible from a good engineering standpoint. So in your quest to outsmart the system and get more for less...well....you aren't.

Originally Posted by Pizzazz

Will this work with the Clark pump? Most likely yes. If the product recovery is around 8-10% as stated here, you can easily double this with just adding a second membrane.

My guess is that the efficiency will go down only a little, so you will benefit from the second membrane.

Absolutely incorrect Statement
(I know Tellie...I should just shut up and have a beer...but...)
It's incorrect because you are only half understanding the design criteria. It's not just recovery rate that matters...but making sure the second membrane would have enough flow rate to not have sky HIGH TDS.
I'm about to give up...but one more....

Originally Posted by Pizzazz

Someone mentioned 320 ppm output - you should consider changing the membrane, I aim to replace membranes when I go above 200 ppm. .

I'm not trying to pick on you Pizzazz but that's just crazy talk. The World Health Organization recognizes 500ppm at the threshold for good clean safe drinking water an the US Military uses 750ppm. 200ppm is lower than most people's domestic water taps and there is no municipality or water maker company that would set such a low threashold. Making such statement will needlessly panic users of water makers and have them throwing out perfectly good membranes....they heard it on the internet right...

One last point.
It's easy for anonymous folks with no "risk" to their credibility or reputation to make guesses and "I think statements" on the internet. Heck it's even fun to think out loud and postulate. But every time I post online about water makers, since that is what I do, I'm putting my credibility on the line and so is Tellie. So why in the world would we say something that we either don't know with 100% certainty to be true or that someone could prove to be wrong? It's what we do for a living and being right about what we do is important to our credibility and reputation. It's much easier to just sit back and think "well that will hurt" and know it won't work and not take the slings or arrows for correcting mistakes...but as a member of the Cruising Community I also don't like to see people waste their hard earned cruising dollars on Chat Room Myths and rumors. I would rather try to stop them before they spread like wild fire and then I'm fielding calls from Spectra owners wanting to buy a second pressure vessel and membrane because they heard it on the internet....

[Pizzazz]

You and I are actually in agreement but may be I should be more precise in my wording. In the example that you give, you start with
0.5 gpm flow and 8 gph product or 30 gph flow, 8 gph product. This is equivalent to 27% recovery which is really close to the max figure of 30% I gave. Since you are already close to the max, clearly adding a second membrane is not possible and the quality of the water will suffer. No argument there, you are actually confirming that you can run the membrane at up to 30% recovery which is what I said.

My point was that the optimum efficiency of the membrane is 8% as quoted by the manufacturer Dow, and if the Spectra is designed to recover only 8-10% product, you can easily increase that to 20-30% with acceptable quality. I do not see anything inaccurate here, it is just math on the vendor data (and my watermaker which is a 0.5 gpm flow type).

For reference, your other sizes imply the following recoveries:
SM12 - 25%
SM30 - 31%
SM40 - 29%
SM50 - 20%
SM60 - 24%

Didn't I say in my post that you can recover 20-30% product? Your own products do that and this is why I like your designs.

Now consider the Spectra Cape Horn Extreme 330, from their manual. Feed flow (on high) = 2.8 gpm, product flow 15 gph which works out to 9% recovery. The reason they are using 9% recovery vs. your design of 20-30% is that they go for max efficiency of the membrane which is possible via the Clark intensifier. So, if you add a second and even a third membrane to the Cape Horn it will work OK in terms of water quality but the efficiency will go down from 1.1A to something higher (either that or the intensifier will fail to maintain 800 psi pressure which is why I said that as long as you can maintain the pressure you are good to add a second membrane). In a way, it would be similar to your SM40 design, it will be much more efficient but not close to the original design goal of 1.1A. If the original poster wants to make more water at slightly higher amperage, let him do it. It makes sense.

I understand that a blanket statement that you can always increase output by adding a membrane threw you off but we were talking about Spectra not your watermakers. I believe my statements are supported by facts, both your figures and Dow's and are true, this time at least. I do not understand your emotion here.

By the way, while it is true that there are a lot of factors, the recovery rate is the only one relevant to our discussion as my analysis above shows. If you go above 30-35% recovery, the salt rejection (in the last membrane) falls off the charts. So, your second accusation is not valid either.

Regarding the 750 ppm being acceptable... I will let that pass. Not worth it.

SV Pizzazz